India in the Persianate Age: 1000–1765

India in the Persianate Age

For family, friends old and new, villagers, farm-hands, musicians, and in memory of fields of golden grain, horses, cows, and dogs (Sadie and Cam), at 100 Lovers Lane, Bainbridge

RICHARD M. EATON

India in the Persianate Age

1000–1765

ALLEN LANE

an imprint of

PENGUIN BOOKS

ALLEN LANE

UK | USA | Canada | Ireland | Australia

India | New Zealand | South Africa

Allen Lane is part of the Penguin Random House group of companies whose addresses can be found at global.penguinrandomhouse.com

First published 2019

001

Copyright © Richard M. Eaton, 2019

The moral right of the author has been asserted

Set in 10.5/14 pt Sabon LT Std

Typeset by Jouve (UK), Milton Keynes

Printed and bound in Great Britain by Clays Ltd, Elcograf S.p.A.

A CIP catalogue record for this book is available from the British Library

ISBN: 978–0–713–99582–4

www.greenpenguin.co.uk

Penguin Random House is committed to a sustainable future for our business, our readers and our planet book is made from Forest Stewardship Council® certified paper.

Contents

List of Illustrations

List of Maps

Acknowledgements

Introduction

Stereotypes and Challenges

Two Transregional Worlds: Sanskrit and Persianate

1 The Growth of Turkic Power, 1000–1300

A Tale of Two Raids: 1022, 1025

Political Culture in the Sanskrit World

Political Culture in the Persianate World

The Ghurid Conquest of North India, 1192–1206

The Delhi Sultanate under the Mamluks, or Slave Kings

Conclusion

2 The Diffusion of Sultanate Systems, 1200–1400

Imperial Expansion Across the Vindhyas

Settlers, Shaikhs and the Diffusion of Sultanate Institutions

The Early Bengal Sultanate

Sultanates of the Deccan: the Bahmanis and Vijayanagara

The Early Kashmir Sultanate

The Decline of the Tughluq Empire

Conclusion

3 Timur’s Invasion and Legacy, 1400–1550

Overview

Upper India

Bengal

Kashmir

Gujarat

Malwa

Emerging Identities: the Idea of ‘Rajput’

Writing in Vernacular Languages

Conclusion

4 The Deccan and the South, 1400–1650

Links to the Persianate World

Successors to the Bahmani State

Political and Cultural Evolution at Vijayanagara

Gunpowder Technology in the Deccan

Cultural Production in the Gunpowder Age

Vijayanagara’s Successors and South India

Conclusion

5 The Consolidation of Mughal Rule, 1526–1605

Overview

Babur

Humayun

Akbar’s Early Years

Emerging Identities: Rajputs

Mughal Expansion Under Akbar

Akbar’s Religious Ideas

Conclusion

6 India under Jahangir and Shah Jahan, 1605–1658

Jahangir

The View from the Frontier

The Deccan: Africans and Marathas

Emerging Identities: the Idea of ‘Sikh’

Assessing Jahangir

Shah Jahan

Conclusion

7 Aurangzeb – from Prince to Emperor ‘Alamgir, 1618–1707

Prince Aurangzeb – Four Vignettes

War of Succession, 1657–9

‘Alamgir’s Early Reign

Emerging Identities: the Marathas from Shahji to Tarabai

‘One Pomegranate to Serve a Hundred Sick Men’

Religion and Sovereignty Under ‘Alamgir

Conclusion

8 Eighteenth century Transitions

Political Changes, 1707–48

Maratha Uprisings

Sikh Uprisings

Emerging Identities: Muslims in Bengal and Punjab

Early Modern Globalization

Conclusion

Conclusion and Epilogue

India in the Persianate World

The Mughals in the Sanskrit World

The Lotus and the Lion

Towards Modernity

Notes

Index

List of Illustrations

1. Bronze image of Śiva as Nataraja, Pala dynasty ( c .750–1161). Śiva temple, Śri Amirthakadeśvara, Mela Kadambur, South Arcot, Chidambaram. Seized from Bengal by Rajendra Chola, c .1022. (© Institut français d’Indologie de Pondichéry).

2. Silver coin of Ajaya Raja II Chauhan (r. c .1110–35) (© Anthony A.J. Hilgevoord and World of Coins). Gold coin of Muhammad Ghuri (r. 1192–1206) (© CoinIndia.com).

3. Warangal fort: interior of Tughluq audience hall, the so-called ‘Khush Mahal’ (1323) (author).

4. Vijayanagara (modern Hampi): northern gateway to the Virupaksha Temple (established twelfth century) (© dbimages/Alamy Stock Photo).

5. Emperor Timur (d. 1405) enthroned. Pen and wash in Indian ink on Japanese paper, by Rembrandt van Rijn (1606–69) (Louvre, Paris, France/Bridgeman Images).

6. Bidar: the madrasa of Mahmud Gawan (completed 1472) (Anne Feldhaus).

7. Shahr-i Sabz, Uzbekistan: the Aq Saray, palace of Timur (completed 1396) (© dbimages/Alamy Stock Photo).

8. Yadgir: wrought-iron cannon on a hill in the centre of the fort (late 1550s) (author).

9. Calicut: Mishkal Mosque (said to date to the fourteenth century, subsequently renovated) (© Hemis/Alamy Stock Photo).

10. Raichur: Krishna Raya with female attendants, depicted on a frieze in the Naurangi Darwaza’s inner courtyard (1520) (author).

11. Detail of Surjan Singh submitting to Akbar, from a painting by Mukund in the Akbar-nama ( c. 1595) (© V&A Museum, London).

12. Vrindavan: interior of the Govinda Deva Temple (completed 1590) (© Robyn Beeche).

13. Nur Jahan loading a musket. Painting attributed to Abu’l-Hasan (1620) (© Raza Library, Rampur, India).

14. Jahangir taking aim at the head of Malik Ambar. Painted by Abu’l-Hasan (1616) (© Chester Beatty Library, Dublin).

15. Amritsar: Harmandir, or Golden Temple (1589) (© Steve Allen/ Getty Images).

16. Jahangir conversing with Jadrup. Painted by Govardhan ( c. 1616–20) (© The Picture Art Collection/Alamy Stock Photo).

17. Prince Aurangzeb attacking a raging elephant. Painting in the Padshah-nama (1633) (© Royal Collection Trust).

18. Lahore: Badshahi Mosque (1671–73) (© DEA / W. BUSS / Contributor/ Getty Images).

19. Shah Jahan enthroned, with his son Dara Shukoh. Watercolour and gold on paper, by Govardhan ( c. 1630–40) (© The History Collection/Alamy Stock Photo).

20. Prince Aurangzeb, probably painted while governor of the Deccan (1653–57) (© Ashmolean Museum, Oxford).

21. Kolhapur, Kavala Naka Square: equestrian statue of Tarabai (erected 1981) (author).

List of Maps

1. The Persianate world, 900–1900

2. Major Indian dynasties, 975–1200

3. Ghurids and Mamluks, 1175–1290

4. India in the time of Muhammad bin Tughluq, 1325–51

5. Fifteenth-century north India

6. Peninsular India, 1565

7. The Mughal empire to the mid seventeenth century

8. India, 1680–1707

Acknowledgements

Books of this sort germinate, grow and ripen in mysterious ways. Yet one thing is clear. Over years of teaching, reading, and immersing myself in fieldwork in India, Pakistan and Bangladesh, many colleagues and friends, perhaps unbeknown to them, helped shape my thinking about South Asian history generally, and about this book in particular. Those to whom I am especially indebted include Said Arjomand, Akeel Bilgrami, Jerry Brennig, Carl Ernst, Munis Faruqui, Barry Flood, Stewart Gordon, Najaf Haider, Harbans Mukhia, Laura Parodi, Shelly Pollock, Ajay Rao, Rajat Ray, Ronit Ricci, Francis Robinson, Samira Sheikh, David Shulman, Ramya Sreenivasan, Audrey Truschke, Phil Wagoner, and André Wink. I also wish to thank those who offered useful comments at talks I gave at the University of California at Berkeley, Vanderbilt University, Jahangirnagar University, Cambridge University’s Trinity College, St Andrews University, and Yale University. For sharpening my focus on what this book is really about, I owe special thanks to Özlem Ayse Özgür. I alone, however, am responsible for any mistakes that might remain.

For his sound advice and steadfast support, I am deeply grateful to Simon Winder and his talented team at Penguin. Linden Lawson’s meticulous copyediting proved immensely valuable. In different ways, many others helped me in this project. For their warm hospitality in Hyderabad (Deccan), I thank V. K. Bawa and Frauke Quader; in Bangalore, Amar Kumar; and in Dhaka, Perween Hasan. I thank Gail Bernstein for commenting on early drafts, Lois Kain for her expert work in preparing the maps, and the staffs at the University of Arizona Main Library and the Chillicothe & Ross County Public Library for their able assistance. I am also grateful to my dean, J. P. Jones, for generously granting me sufficient leave time to complete this project.

That might never have happened, however, were it not for the tranquility and the fellowship of my sister Beth’s lovely farm in the Appalachian foothills. The following chapters were drafted in the walnut-panelled hallway of her rambling, nineteenth-century farmhouse, surrounded by fields of corn and soybeans. It was a delightful place to work – inscribed on the brick and stone hearth was an old Scottish adage: ‘East, West, Hame’s Best’. On occasion, I might spot an eagle soaring high in the skies outside, or hear the sound of horses galloping along the pasture fence that lay only a few meters beyond the hallway’s big front door. I dedicate this book to those treasured days.

India in the Persianate Age

Introduction

STEREOTYPES AND CHALLENGES

This book focuses on South Asian history from roughly 1000 to the late eighteenth century, one of the most compelling, consequential and controversial periods in India’s long history. While providing a broad overview of the subcontinent during this period, the book also seeks to challenge lingering stereotypes that have taken hold in recent decades.

One such stereotype is that India had remained a largely stagnant civilization until stimulated by European rule in the eighteenth century. In contrast, the current volume paints a picture of India’s repeated self-transformation during these eight centuries. It was between the eleventh and eighteenth centuries, after all, that India witnessed, among other things, the disappearance of Buddhism, the appearance of the Sikh religion, the growth of the world’s largest Muslim society, the transformation of vast tracts of land from jungle to fields of grain and the integration of tribal clans into the Hindu social order as castes. This era also witnessed India’s emergence as the world’s industrial powerhouse, based on the export of manufactured textiles. The notion that India merely stood still for eight centuries is, to say the least, mistaken.

Another stereotype addressed in this volume is the notion of India as a self-contained and territorially bounded essence, historically isolated from outside. Rather, this book stresses South Asia’s contacts with the societies and cultures of Central Asia, Africa, East Asia, South-east Asia and, especially, the Middle East. In fact, most of the historical changes mentioned above cannot be understood without situating India in the context of its relations with neighbouring peoples.

A third and related stereotype is that of India as an essentially self-generated Hindu and Sanskritic civilization that evolved on its own, rather than a hybridized composite produced from protracted interaction with other peoples and cultures. The present volume affords an excellent opportunity to examine this theme since its chronological scope covers the period of South Asia’s intense contact with other regions, particularly with the Iranian plateau, with Persian culture and with Islam. Indeed, the period extending from c.1000 to c.1800 is conventionally referred to as India’s ‘Muslim period’, inaugurated by a ‘Muslim conquest’ of India. But there is good reason to question such characterizations.

Consider an analogous world-historical encounter. By the early sixteenth century Spanish conquistadors had sailed across the Atlantic Ocean, established large empires in Central and South America, planted new colonies there, forcibly uprooted native American religious and political institutions and conducted a vigorous, state-sponsored programme of Christianizing the continent’s native populations, as a result of which the vast majority of the region’s peoples are Roman Catholic today. Yet historians never refer to this great historical moment as a ‘Christian conquest’ of America. Rather, it is conventionally understood as the ‘Spanish conquest’. But generations of historians have referred to the equally momentous events that took place in India towards the end of the twelfth century not as a Turkish conquest but as ‘the Muslim conquest’, even though the Sanskrit term typically used by contemporary Indians to describe the conquerors was not ‘Muslim’ but ‘Turk’ (turushka). Further complicating the idea of some religion-based ‘clash of civilizations’ is the fact that Muslims who had already settled in north India – specifically, in early-thirteenth-century Benares – fought with Indian dynasties against these invading Turks.¹ So a key question that should be asked at the start is: what explains the very different ways in which the American and Indian cases are conventionally characterized? Why is religion foregrounded in one, but not in the other? What hidden assumptions lurk behind our continued use of such different categories when we refer to these otherwise comparable encounters?

Much is at stake in these questions. First, the notion of a ‘Muslim conquest’ may well result from the inappropriate application of present-day understandings of religion to earlier times, as though religions had always been self-contained and closed belief systems, impervious to change over time and making totalizing claims on people’s identities. Then there are political issues. Ever since the end of British imperial rule in 1947, the two largest states in South Asia, India and Pakistan, have remained bitter rivals, with one of them making Islam a state religion and, at least initially, the sole criterion of its national identity. As Pakistan’s President Zia-ul-Haq stated in 1981, ‘Take Islam out of Pakistan and make it a secular state; it would collapse.’² It is, then, hardly surprising that the three wars fought by these now nuclear states have only reinforced the notion of religion as the primary force that had ‘always’ divided South Asia’s inhabitants.

Of course, the idea of cultural alterity, or ‘otherness’, long predated the creation of the two states. Think of the figures of speech found in two very different literary traditions, the Persian and the Sanskrit. From the mid eleventh century a dynasty of Central Asian Turks, the Ghaznavids, ruled over much of the Punjab from Lahore. However, their Indian rivals identified them in their inscriptions and texts not as Muslims but as the ‘Lords of the Horses’, an apparent reference to their dependence on cavalry warfare and their control over trade routes leading to Central Asia, a major source of war-horses. That is, these Turks were understood as powerful but familiar rivals in north India’s crowded stage of contending lineages. But then in the late twelfth century another Turkish group, the Ghurids, would sweep away both the Ghaznavids and north India’s martial clans, later called Rajputs, and put them on a path to eventually establishing the Delhi sultanate (1206–1526), a sultanate being, of course, a kingdom ruled by a sultan. Unlike the Ghaznavids, however, these later Turks had not yet been assimilated as one of north India’s many ruling houses: on the contrary, they were wholly alien and unfamiliar, not to mention destructive. Accordingly, a contemporary Sanskrit epic, the Prthviraj-vijaya, lustily stigmatized them as outright ‘barbarians’ (mlechhas), ‘demon-men’ (nararaksasam), enemies of cows and ‘given to eating foul foods’.³ Yet, as destructive and alien as they were, the Ghurid Turks – like their Ghaznavid predecessors – were not identified by their religion. As the historian Cynthia Talbot notes, the image of Muslims in contemporary Indian texts ‘oscillated depending on the prevailing political conditions: in times of military conflict and radically fluctuating spheres of influence, the rhetoric was often negative in tone; whereas long-established Muslim rulers were conceptually assimilated into the Sanskritic political imagination’.⁴

That said, the authors of the Persian chronicles, unlike their Indian counterparts, certainly did see the world through the lens of religion: people were either Muslim believers or infidels. But, we must ask, for whom did these writers speak? It is one thing for a pious chronicler to colour an event in ways that conformed to – or violated – his own sense of a properly ordered world. However, how culturally different communities actually interacted with one another, or what sorts of political and social modi vivendi they reached, can be another thing altogether. This means that, while Persian chronicles are indispensable in reconstructing India’s history in our period, it would be a mistake to rely on that genre alone. Hence the present volume parts company with British-period historians of India, who obsessively adhered to written data to the exclusion of other kinds of evidence and placed excessive trust in Persian chronicles, which for them formed an unshakeable basis for the reconstruction of India’s post-eleventh-century past. Not surprisingly, British histories of India written during the Raj tended to reproduce the very believer-vs-infidel mindset of the chroniclers whose Persian texts they used.

Another reason why many nineteenth- and early-twentieth-century historians replicated the religiously defined worldview of medieval Persian chroniclers relates to Britain’s rationale for occupying India. The British came to justify the Raj on the grounds that they had introduced India to an enlightened era of sound and just government, a position logically requiring that rulers immediately preceding them be construed as despotic and unjust. Perhaps the clearest case of history-writing in service of the Raj is the work of Sir Henry M. Elliot, whose translations of Indo-Persian chronicles, Bibliographical Index to the Historians of Muhammedan India, first appeared in 1850. Elliot sought to use such chronicles to show readers how destructive Muslim rulers had been before the arrival of British rule. As he wrote in the Preface:

The few glimpses . . . we have of Hindus slain for disputing with Muhammadans, of general prohibitions against processions, worship, and ablutions, and of other intolerant measures, of idols mutilated, of temples razed, of forcible conversions and marriages, of proscriptions and confiscations, of murders and massacres, and of the sensuality and drunkenness of the tyrants who enjoined them, show us that this picture is not overcharged.⁵

Elliot presents the advent of European rule, by contrast, as a period ‘when a more stirring and eventful era of India’s History commences; and when the full light of European truth and discernment begins to shed its beams upon the obscurity of the past’. Therefore, he concludes, reading translations of Indo-Persian chronicles – which he characterized as dull, prejudiced, ignorant and superficial – ‘will make our native subjects more sensible of the immense advantages accruing to them under the mildness and equity of our rule’.⁶ Within seven years, India would be consumed by the horrific Revolt of 1857 and its brutal suppression by British troops. Nonetheless, the rhetoric of the Raj’s ‘mildness and equity’, in contrast to the ‘Muhammadan’ tyranny said to have preceded it, would prevail throughout Britain’s occupation of India.

Other factors also inclined the British to see Indian history through the lens of religion. Students of South Asian history are aware of the charge that European rulers had deployed classic ‘divide-and-rule’ measures as a strategy for governing India. Already in the late eighteenth century, as the East India Company was gaining a political toehold on parts of South Asia, Governor General Warren Hastings established a legal system in which Muslims and non-Muslims were tried by separate law codes; henceforth, Muslims and non-Muslims would constitute juridically separate communities. The British then went on to establish a formidable array of publications – decennial census reports, district gazetteers, ethnographic surveys, etc. – that pigeonholed Indians into separate, watertight compartments using religion as a principal category. All this, so the argument goes, had the insidious effect of enhancing – some would even say creating – cultural divisions in an otherwise relatively harmonious Indian society.

Consider, too, how religion dominated European notions of Indian time. Comprehensive histories of India published in the nineteenth and twentieth centuries were typically divided into three periods: ancient, medieval and modern. The scheme dates at least to the 1817 publication of James Mill’s The History of British India, which divided India’s history into Hindu, ‘Mahomedan’ and British eras.⁷ This tidy, tripartite scheme was actually a transposition on to India of the same ancient–medieval–modern scheme by which, ever since the Renaissance, Europeans had periodized their own history. In the South Asian case, however, those three temporal units were made to correspond to three culturally defined and supposedly homogeneous communities that had successively ruled most of the subcontinent. Formulated in this way, the system posited two great ruptures in Indian time. The first, which defined the transition from ‘ancient’ to ‘medieval’, implied a descent from an earlier Hindu ‘golden’ age to one of ‘Mahomedan’ tyranny. To India’s British rulers, this decline corresponded to Europe’s fall from an earlier age of Greco-Roman splendour to its own medieval period, initiated by the so-called ‘Dark Ages’. Implicitly, then, the appearance of Muslim Turks in India was analogous to that of the Visigoths or Vandals in Rome: all were construed as alien outsiders whose armed intrusions had violated a sacred realm. Such a formulation allowed British imperialists to imagine India’s second great historical rupture – the transition from ‘medieval’ times (i.e. Muslim rule) to modernity (i.e. British rule) – as having validated the coming of European governance as a blessing for a benighted land. By this self-serving formulation, Britain had liberated India from eight centuries of ‘Muhammadan’ stagnation.

While Indian Muslims in the modern period certainly did not share this view of India’s middle period, many did see the advent of Islam as a transformative moment in India’s history. Early leaders of the Pakistan movement, seeking a historical basis for justifying the creation of a separate Muslim state in post-British South Asia, propounded the so-called ‘two-nation’ theory. According to this understanding, India’s Muslims had comprised a homogeneous and self-aware community objectively distinct from India’s non-Muslims ever since the eighth century, when the earliest known Muslim community had appeared in the region. Therefore, the creation of an Islamic state merely acknowledged constitutionally what was held to have been a social reality for over a thousand years. In this way, too, Muhammad bin Qasim, the eighth-century Arab conqueror of Sind, in today’s Pakistan, could be conjured up as a proto-nationalist figure, even as Pakistan’s ‘first citizen’.⁸

Conversely, in their efforts to locate their own moments of glory in India’s past, many Hindu nationalists of the first half of the twentieth century imagined rebels against pre-colonial ‘Muslim’ states as heroes who were, in some small or inchoate way, struggling on behalf of an India-wide, pan-Hindu collectivity. Thus in the early twentieth century, during the twilight years of the Raj, two opposing nationalist narratives emerged, both driven by religion. And since any form of nationalism selectively picks and chooses from its past in order to endow the present with meaning, if not inevitability, both Hindus and Muslims politicized South Asia’s history, in particular the eight centuries prior to the British arrival. One community’s heroes became the other’s villains, and vice versa, while both narratives interpreted the past in order to explain the present and justify an imagined future. India’s ‘medieval’ history, in short, became a political football.

Although British rulers, Indian nationalists and Muslim separatists were motivated by very different agendas, each understood India’s middle period as one in which religion comprised the fundamental building block of community identity, with the Muslim presence in India looming especially large in South Asia’s collective consciousness. This is clearly reflected in the tradition of history-writing since the nineteenth century. In book after book, the tendency has been to list events, kings, battles and literary and religious texts in chronological order, each of them neatly divided into separate Hindu and Muslim compartments.⁹ India was thus given two Procrustean beds, one Muslim and one Hindu, into one or the other of which nearly everything had to fit – architecture, dress, art, literature, language, and so on.¹⁰ The British art historian Percy Brown, for example, could publish a two-volume study on Indian architecture, one volume covering the ‘Buddhist and Hindu periods’ and another the ‘Muslim period.’

The reading of history in terms of mutually exclusive religions has, however, come at enormous cost. For one thing it has made it difficult to account for, or even to see, larger cultural processes. Consider the earliest genre of Hindi literature – the so-called premakhyans, or Sufi romances – which appeared in the eastern Gangetic plain between the fourteenth and sixteenth centuries. This literature was composed originally in the Persian script by Sufis who narrated the seeker’s mystical quest for union with God, but it did so using characters who were ostensibly Hindu in name and cultural/religious practice, in a landscape saturated with Indian deities, mythology, flora and fauna. Failing to fall neatly into either Hindu or Muslim categories, this literary genre baffled Ramchandra Shukla and other early-twentieth-century nationalist writers, who engaged in long and fruitless debates over whether or not this literature was truly Indian.¹¹

The convention of seeing India only in terms of religion, and of dividing its history into three religiously defined units of time, is thus well entrenched. Although the present volume covers what historians in the tradition of James Mill labelled the ‘Muslim’ age, the aim is nonetheless to narrate this period on its own terms, and not to project on to it today’s values or biases. For not only did India’s socio-cultural landscape differ vastly from that of today: the conceptual categories by which peoples of earlier times understood that landscape did too. We might start, then, by rethinking the notion of a ‘Muslim’ conquest and, indeed, the proper place of religion in India’s history during this period. But if religion is not to serve as the key to India’s past, what might?

TWO TRANSREGIONAL WORLDS: SANSKRIT AND PERSIANATE

Western Civilization, Dar al-Islam (‘the abode of Islam’), Christendom, the Motherland, the Free World, the Promised Land, the Third World, the Middle Kingdom – these are just some of the terms in which people have imagined geographical space, attempting in each instance to impose culture or ideology on to territory. It can be a vexed enterprise. In recent years the Sanskritist and historian Sheldon Pollock, suggesting a very different way of thinking about cultural space, coined the term ‘Sanskrit cosmopolis’, referring to the diffusion of Indian culture across a vast swathe of Southern Asia between the fourth century and the fourteenth.¹² Sanskrit place names alone attest to the geographical sweep of a culturally connected zone between Afghanistan’s Kandahar (Skt Gandhara) and the South-east Asian city state of Singapore (Skt Singhapura).

For Pollock, what characterized this Sanskrit world was not religion but the ideas elaborated in the entire corpus of Sanskrit texts that, between the fourth and fourteenth centuries, circulated above and across the world of vernacular, regional tongues. Sanskrit, like only a few others, was a language that travelled: it was not a ‘language of place’. Not being identified with a particular ethnic or linguistic group or with a particular region, Sanskrit was transregional by nature, or, as Pollock puts it, ‘a language of the gods in the world of men’. Texts composed in Sanskrit embraced everything from rules of grammar to styles of kingship, architecture, proper comportment, the goals of life, the regulation of society, the acquisition of power and wealth, and much more. The circulation of these texts and of the people who carried them created a network of shared idioms and styles that made similar claims about aesthetics, polity, kingly virtue, learning and the universality of dominion. Fundamentally, the Sanskrit world – that is, the vast sweep of territory in which such texts circulated and were considered normative – was concerned with defining and preserving moral and social order.

Moreover, this cultural formation expanded over much of Asia not by force of arms but by emulation, and without any governing centre or fortified frontiers. It was thus comparable to the Hellenized world that embraced the Mediterranean basin and the Middle East after Alexander the Great. For that world, too, was a cultural zone without political borders, in which people of many ethnic or religious backgrounds readily subscribed to the prestige of Greek language, sculpture, drama, cuisine, architecture and so on, but without paying taxes to a Greek official or submitting to the might of Greek soldiers. We may contrast this ‘cosmopolis’ idea with any classical empire, such as the Roman, with its centralized governing structure, sharp distinction between citizens and non-citizens, fortified frontiers and reliance on the hard power of coercive force as opposed to the soft power of models that encourage emulation.

The Sanskrit world that Pollock describes was, however, only one such formation to have appeared in South Asian history. From about the eleventh to the nineteenth centuries a similar, Persianate world embraced much of West, Central and South Asia. Both expanded and flourished well beyond the land of their origin, giving them a transregional, ‘placeless’ quality. Both were grounded in a prestige language and literature that conferred elite status on its users. Both articulated a model of worldly power – specifically, universal dominion. And while both elaborated, discussed and critiqued religious traditions, neither was grounded in a religion, but rather transcended the claims of any of them. Decoupled from particular religious systems, both of these transregional traditions could and did spread over great expanses of territory, and were embraced by peoples of varied ethnic and religious backgrounds. Fundamentally – and this is the underlying theme of this book – much of India’s history between 1000 and 1800 can be understood in terms of the prolonged and multifaceted interaction between the Sanskrit and Persianate worlds.

The Persianate world, 900–1900

But what exactly is the Persianate world, and how did it evolve?¹³ Several centuries after the Arab conquest of the Iranian plateau in the seventh century, Persian-speakers gradually recovered a rich but largely submerged pre-Islamic Persian civilization. The linguistic dimension of this movement saw the emergence of New Persian – a hybrid of the indigenous Middle Persian of Iran’s Sasanian period (AD 224–651) and the Arabic brought to the Iranian plateau in the course of the Arab conquest. This new language appeared first in spoken form across the Iranian plateau and deep into Central Asia. A written form using a modified Arabic script emerged in the ninth and tenth centuries, when Persian writers in present-day north-eastern Iran, western Afghanistan and Central Asia began appropriating the cultural heritage of both Arab Islam and pre-Islamic Iran.

Initially, at least, these developments were promoted and patronized by the court of the Samanid kings in Central Asia (819–999). Based in Bukhara (in today’s southern Uzbekistan), the Samanid domain straddled major trade routes connecting the Iranian plateau with the Mediterranean to the west, India to the south and, via the Silk Road, China to the east. Bukhara thus lay in a commercially vibrant zone. It was also multilingual, as Arabic and Turkish were both commonly used there, as was, until the eleventh century, Sogdian. But New Persian (henceforth simply ‘Persian’) was now the lingua franca, having replaced the region’s indigenous Iranian languages and dialects.

As with the Sanskrit texts, from the eleventh century onwards a large corpus of imaginative literature in Persian began to circulate widely through West, Central and South Asia. A case in point is the cycle of epics based on the historical Alexander penned by such luminaries as Firdausi (d. 1020) in Iran, Nizami (d. 1209) in Georgia, Amir Khusrau (d. 1325) in India and Jami (d. 1492) in Afghanistan. Although composed a great distance apart, and circulating over an even wider one that spanned many vernacular cultures, these works enabled diverse peoples to imagine and inhabit a single cosmopolitan space enlivened by Alexander’s real – or imagined – exploits.¹⁴ Such works of literature helped knit together a ‘Persianate world’ across West, Central and South Asia. However, like Sanskrit texts, Persian literature had no single geographical or political centre, especially after the thirteenth century when Mongol invaders overran Central Asia and northern Iran, destabilizing their courts. From that point on, peoples in far-flung regions such as the Caucasus or India might retain everyday use of their local languages while cultivating, and even producing, great works of Persian literature.¹⁵ By the fourteenth century Persian had become a vibrant and prestigious literary language, a widely used medium in state bureaucracies, and the principal contact tongue for inter-regional diplomacy along the Silk Road between Anatolia and East Asia. In Mongol-dominated China, it served not only as a lingua franca but as the official foreign language. The Venetian merchant-traveller Marco Polo (d. 1324) mainly used Persian in China, as he did, in fact, throughout his travels on the Silk Road. So did his near-contemporary and even greater globetrotter Ibn Battuta (d. 1377), who travelled many of the same pan-Asian circuits in fourteenth-century Asia.¹⁶

Of particular relevance for understanding India’s changed political order after the late twelfth century is what Persian writers had to say about power and authority. Crucially, the same culturally diverse milieu that had nurtured the literary and bureaucratic use of Persian under Samanid patronage also shaped a particular conception of a universal ruler or ‘sultan’, the title preferred by such men throughout the Persianate world. Occupying a political space above all ethnic groups and religious communities, this figure was understood as both universal and supreme: he occupied unlimited sovereign space and commanded the loyalty of all lesser political actors. The crystallization of the idea of the sultan in the tenth and eleventh centuries resulted from two factors in particular: the steady decline of the Abbasid caliphate in Baghdad, which in theory ruled over the entire eastern Islamic world, including Central Asia; and the infiltration of waves of Turkish-speakers from eastern Asia into urbanized Central Asia and northern Iran. Some came as military recruits, others as pastoral nomadic migrants, others as powerful confederations of warriors. To accommodate these new realities, political thought in South-west Asia underwent drastic revisions. In particular, spiritual and political authority split into separate spheres, with the caliph retaining his religious authority and the sultan exercising effective political power. Making the best of a bad situation, a leading theologian of the time, Abu Hamid al-Ghazali (d. 1111), pronounced that any government was lawful so long as its ruler, or sultan, acknowledged the caliph’s authority in spiritual matters. Reciprocally, caliphs accepted the secular authority of upstart sultans under the fiction of having appointed them to their office.

Stoking memories of pre-Islamic Iran, poets and chroniclers endowed these sultans with the same pretentions to absolutist rule as pre-Islamic Persian emperors. In the early twelfth century the historian Ibn Balkhi conceived of kingship in that earlier age as based on the supreme principle of justice, for, he wrote, every king had taught his heir apparent the following maxim:

There is no kingdom without an army, no army without wealth, no wealth without material prosperity, and no material prosperity without justice.¹⁷

Persian scholars such as Ibn Balkhi made no attempt to yoke state power to Islam or to any other religious tradition; instead, it was justice that bound their world together. Notably, long before Renaissance or Enlightenment thinkers in Europe began theorizing the separation of Church and State, intellectuals in eleventh- and twelfth-century Iran and Central Asia were already doing precisely that. Such a secularist conception of government would have far-reaching implications for rulers styling themselves sultans in areas as ethnically diverse as India. In fact, by the time it reached India, the term ‘sultan’ had become so detached from ethnicity or religion that Hindu rulers, aspiring to the most powerful titles then available to them, adopted it. In 1347 Marappa, one of the founders of the Deccan kingdom of Vijayanagara, declared himself ‘sultan among Indian kings’ (hindu-raya-suratalah), a title used also by his earliest successors.¹⁸

India’s eventual inclusion in this expanding Persianate world was thus facilitated by, among other things, a ruling ideology that had co-opted the political authority of a caliph, embraced the principle of universal justice and accommodated cultural diversity. Such an inclusivist political ideology happened to be well suited for governing a north Indian society that was itself extraordinarily diverse religiously, linguistically and socially. Moreover, the elevation of justice, not religion, as the measure of proper governance allowed Persianized states to flourish throughout India, notwithstanding a ruler’s own religion. As argued by Ziya al-Din Barani (d. c. 1357), a leading historian and theorist of the early Delhi sultanate, whereas any country could flourish under a non-Muslim ruler as long he was just, no country ruled by a Muslim would flourish if he was unjust.¹⁹

What is perhaps most remarkable about the Persianate world, however, is how readily its core ideas diffused not only within Indian territories governed by Persianized states such as the Delhi sultanate, but also into territories lying beyond such states. A distinctively Persianate ideology privileging the notion of justice and connecting economy, morality and politics infiltrated peninsular India even while that region was governed by independent Hindu rulers. At some point in the twelfth or thirteenth centuries the Telugu poet Baddena, writing at the Kakatiya court at Warangal in southern India, penned these striking lines:

To acquire wealth: make the people prosper. To make the people prosper: justice is the means. O Kirti Narayana! They say that justice is the treasury of kings.²⁰

In linking wealth, prosperity and justice, this terse aphorism seems to paraphrase Ibn Balkhi. Moreover, it is clear that the idea of justice, so central to Persian political thought, had been freely borrowed by Baddena and not imposed from without.²¹ Like Mongol rulers on the Iranian plateau, Baddena had assimilated a Persianate vision of political and moral order, even though he lived very far from the Delhi sultanate.

Apart from political ideology, other aspects of Persian culture spread throughout South Asia after the thirteenth-century Turkish conquest of north India, including styles of architecture, dress, music, courtly comportment, cuisine and, especially, vocabulary. As the geographical reach of Persian letters expanded, so did the production of dictionaries, whose compilers sought to make literature produced in different parts of the Persophone world mutually comprehensible; by the nineteenth century, many more Persian-language dictionaries had been produced in India than in Iran, suggesting how thoroughly India had been absorbed into that world. Indeed, by the fourteenth century Persian had already become the most widely used language for governance across South Asia, as Indians filled the vast revenue and judicial bureaucracies in the Delhi sultanate and its successor states, and later in the Mughal empire (1526–1858) and its successors. As a result, a wide range of Persian words infiltrated the vocabulary of many of South Asia’s major regional languages.

All of which brings us back to the theme of periodization, and the rationale for this book’s chronological borders of 1000 and 1765. Recent generations of historians of India have rightly eschewed the old tripartite Hindu–Muslim–British scheme and have reverted to its European predecessor, the ancient-medieval-modern one. But the precise meaning of these timespans, especially the second, has remained elusive. Instead of giving substance to the term ‘medieval’, historians have produced a host of high-quality regional studies covering the whole or part of the period 1000–1800 – e.g. on Bengal, Gujarat, Malabar, Orissa, the Punjab, the Deccan plateau, the Delhi region, the Tamil country. As a result, the term ‘medieval’ when applied to India as a whole has become something of an orphan – repeatedly invoked, but lacking meaning. As the historian Daud Ali notes:

the category of medieval has gradually been evacuated of any definitive substance in most national historiographies, in favour of a sort of cacophony of regional isolates simply holding the fort until the cavalry arrives²²

By ‘the cavalry’ Ali appears to mean some new conceptual handle or idea that might confer meaning on the term ‘medieval’, other than that of a religiously defined Muslim era.

I argue that there is such a handle. As it happens, the period of India’s history conventionally labelled ‘medieval’ coincides with the eastward diffusion of Persianate culture across almost all the Indian subcontinent and its interaction with its Sanskrit counterpart. The story of this interaction – the encounter between the Persian and Sanskrit worlds – is both rich and complex. It is the subject of this book.

1

The Growth of Turkic Power, 1000–1300

A TALE OF TWO RAIDS: 1022, 1025

In the early second millennium, within only three years of each other, two armies marching from opposite directions raided north India. Neither would remain to govern or colonize conquered territory. Although both expeditions were successful in their own ways, they harboured different objectives, had different martial traditions and were informed by very different political systems. The two invasions also suggest why the opening of the second millennium marked a major transition in India’s long history.

The first of the two raids was led by a general acting on the authority of Rajendra I (r. 1014–44), maharaja of the Chola empire (848–1279) towards the extreme southern end of the Indian peninsula. In 1022 his army marched 1,600 kilometres north from the Cholas’ royal and ceremonial capital of Tanjavur, moving along India’s eastern coast. After subduing kings in Orissa, Chola warriors defeated rulers in the western and the south-central districts of the Ganges delta. Then, in a fiercely fought pitched battle, they defeated Mahipala, maharaja of the Pala empire (c.750–1161), at the time the dominant power in India’s easternmost region of Bengal. The southerners crowned their victory by carrying off a bronze image of the deity Śiva, which they seized from a royal temple that Mahipala had presumably patronized [see Fig. 1]. In the course of this long campaign, the invaders also took from the Kalinga raja of Orissa images of Bhairava, Bhairavi and Kali. These, together with precious gems looted from the Pala king, were taken down to the Chola capital as war booty.¹

Major Indian dynasties, 975–1200

Before leaving the delta, however, Chola officers directed an operation unusual for military campaigns: they arranged for water from the Ganges River to be collected in pots and carried on the army’s long march back to Tanjavur. In the Godavari River delta, midway between Bengal and the Chola heartland, Rajendra, who had been consolidating Chola rule in coastal districts north of his capital, joined the victorious army. From there, the combined forces triumphantly returned to Tanjavur. The Cholas were at this time nearing the zenith of their might and glory; they would soon become the dominant power in the eastern Indian Ocean, their influence stretching across the Bay of Bengal to Sumatra. In their own estimation, they occupied the centre of the earthly and cosmic worlds.

In October 1025, not long after Rajendra Chola’s return from his conquests in Bengal, the son of a Turkish-speaking Central Asian slave marched out of Ghazni in eastern Afghanistan with 30,000 cavalry behind him. Heading south-east through the craggy ravines of the Sulaiman Mountains, he and his troops descended from the Afghan plateau into the low, lush Indus valley. Like his Chola contemporaries, Sultan Mahmud of Ghazni (r. 997–1030) planned to attack a specific north Indian target – the wealthy temple of Somnath, an important pilgrimage site on the shores of the Arabian Sea on Gujarat’s southern coast. Built of stone a hundred years earlier and situated in a fortress that was surrounded by the sea on three sides, the temple of Somnath, like that of Mahipala in Bengal, was dedicated to the god Śiva. In December 1025, having crossed the Indus and marched across western India’s forbidding Thar Desert, Mahmud reached the site, successfully besieged the fortress and plundered the famous temple of its riches. He also ordered its Śiva image to be broken up and its pieces taken back to Ghazni, his capital, to be set in a floor and walked upon.²

This was not Mahmud’s first raid on north India. The sultan had already launched more than a dozen, beginning with an attack in 1001 on Peshawar, at the foot of the strategic Khyber Pass which connects the Afghan highlands with the Indus valley. Almost annually, similar offensives took place against prominent cities of the Punjab and the upper Ganges valley. In each of them, Mahmud’s men brought plundered wealth back through the mountain passes leading to Ghazni. What distinguished the Somnath raid from the others, however, was the way in which it captured the imagination of Persian chroniclers: those contemporary with the raid hailed Mahmud as an arch-iconoclast, piously responding to Islam’s prohibition against image-worship. Subsequent chroniclers even lionized him as the founder of Turkish rule and of Islamic sovereignty in South Asia, although in fact he was neither of these.

In striking contrast to Persian chronicles, which made so much of Mahmud’s raid on Somnath, Sanskrit inscriptions recorded by local Hindus made no mention of it at all. On the contrary, accounts dating to the months and years after the raid convey a sense of undisturbed business as usual for both the temple and the bustling seaport of Somnath, a major commercial entrepôt that imported war-horses from the Persian Gulf and exported locally produced textiles to markets around the Arabian Sea. Twelve years after the attack, a king from the Goa region recorded performing a pilgrimage to the temple, but he failed to mention Mahmud’s raid. Another inscription dated 1169 mentioned repairs made to the temple owing to normal deterioration, but again without mentioning Mahmud’s raid. In 1216 Somnath’s overlords fortified the temple to protect it not from attacks by invaders from beyond the Khyber Pass, but from those by Hindu rulers in neighbouring Malwa; apparently, such attacks were so frequent as to require precautionary measures.³ The silence of contemporary Hindu sources regarding Mahmud’s raid suggests that in Somnath itself it was either forgotten altogether or viewed as just another unfortunate attack by an outsider, and hence unremarkable.

In fact, the demonization of Mahmud and the portrayal of his raid on Somnath as an assault on Indian religion by Muslim invaders dates only from the early 1840s. In 1842 the British East India Company suffered the annihilation of an entire army of some 16,000 in the First Afghan War (1839–42). Seeking to regain face among their Hindu subjects after this humiliating defeat, the British contrived a bit of self-serving fiction, namely that Mahmud, after sacking the temple of Somnath, carried off a pair of the temple’s gates on his way back to Afghanistan. By ‘discovering’ these fictitious gates in Mahmud’s former capital of Ghazni, and by ‘restoring’ them to their rightful owners in India, British officials hoped to be admired for heroically rectifying what they construed as a heinous wrong that had caused centuries of distress among India’s Hindus. Though intended to win the latters’ gratitude while distracting all Indians from Britain’s catastrophic defeat just beyond the Khyber, this bit of colonial mischief has stoked Hindus’ ill-feeling toward Muslims ever since.⁴ From this point on, Mahmud’s 1025 sacking of Somnath acquired a distinct notoriety, especially in the early twentieth century when nationalist leaders drew on history to identify clear-cut heroes and villains for the purpose of mobilizing political mass movements. By contrast, Rajendra Chola’s raid on Bengal remained largely forgotten outside the Chola country.

On the surface, the military operations of Rajendra Chola and Mahmud of Ghazni would seem to have had much in common: both armies marched some 1,600 kilometres with a view to attacking and plundering specific north Indian sites; neither had any intent of occupation, annexation or permanent government; and both desecrated royal temples, carrying off plundered images to their respective capitals. But the differences between the two invasions are far more important than their similarities, since they highlight radically different political cultures in early-eleventh-century South Asia. The older of these two political cultures, the Chola, was informed by a body of Sanskrit texts that had circulated across India for many centuries before the rise of Chola power in south India. The other culture, informed by an analogous body of Persian texts, had come into being only two centuries before Mahmud’s day.

POLITICAL CULTURE IN THE SANSKRIT WORLD

The culture that had informed Rajendra Chola’s political actions, including his raid on Bengal, was derived from Sanskrit treatises on power, wealth and rulership. It presumed a political universe crowded with little kings, bigger kings and emperors – that is, kings of kings. It also presumed a world of shifting political sands, where rulers had neither permanent enemies nor permanent allies. In conducting the business of warfare, therefore, classical Indian thought recommended that enemies not be annihilated, but rather co-opted and transformed into loyal subordinates⁵ who could be put to use as allies against future enemies. Thus the same inscription that describes the Chola raid on Bengal records that when Rajendra returned from his victorious northern expedition:

He [then] entered his own [capital] town, which by its prosperity despised all the merits of the abode of the gods – his lotus feet [all along] being worshipped by the kings of high birth who had been subdued [by him].⁶

In reality, of the eight kings that Rajendra or his generals are said to have fought on this expedition, one was killed in battle, two others fled the battlefield and the rest were ‘conquered’. These defeated kings – or at least, the five who survived the Chola invasion – were not executed or publicly humiliated. Instead, they became loyal vassals.

Such an outcome conformed to well-defined norms of inter-state politics long canonized in classical Indian thought.⁷ According to these norms, territory was imagined as something like a large chessboard on which kings manoeuvred with allies and against rivals with a view to creating an idealized political space called the Circle of States, or mandala. The term referred to a series of concentric circles, where one’s own capital and heartland was at the centre, surrounded by a second circle of one’s allies, and a third circle of one’s enemies. Beyond that lay a fourth circle occupied by one’s enemies’ enemies, understood as potential allies with whom a king endeavoured to ally himself. With all of India’s major dynastic houses playing by the same geostrategic rules, the result was not only intense political jockeying and perpetual conflict, but overall stalemate and equilibrium.⁸ No single dynastic house could achieve lasting dominance over large tracts of territory within India, much less over South Asia as a whole. Indeed, it is hardly surprising that chess itself originated in India around the sixth century, just when these geopolitical ideas were taking hold.⁹

Rajendra’s father Rajaraja (r. 985–1014) is widely acclaimed as the greatest of Chola emperors, as judged by his conquests and the literature and art he patronized, including his building of one of the grandest temples in south India, the Rajarajeśvaram (or Brihadeśwara) Temple in his capital of Tanjavur. The temple was designed to replicate cosmic space and to situate itself at the centre of that space; one of its names is ‘Daksinameru’, or the southern Mt Meru – that is, the axis of the universe.¹⁰ Much of the wealth necessary for patronizing the king’s cultural projects derived from his successful deployment of the mandala strategy on India’s geopolitical chessboard. From the Cholas’ heartland in the fertile Kaveri delta, Rajaraja had waged a series of victorious military campaigns, defeating in turn the Pandya kings of Madurai to his south and the Cera kings of Kerala to his west. Since these two dynasties had been allied with the Buddhist kings of Sri Lanka, Rajaraja launched a naval expedition to the island kingdom and sacked its ancient capital of Anuradhapura, making him the first Indian king to embark on overseas conquests. Just as importantly, these conquests validated Rajaraja’s claims to being a universal emperor (chakravartin) since, according to classical Indian political thought, such an emperor had to perform a digvijaya, or ‘conquest of the quarters’ – that is, kingdoms to the south, west, north and east.

In 1014, when Rajaraja died, Rajendra, who had been co-emperor at the end of his father’s reign, became Chola emperor in his own right. One of Rajendra’s inscriptions records that he soon thereafter ‘turned his attention to the conquest of the quarters [digvijaya] backed by a powerful army’. In 1017 he launched a fresh invasion of Sri Lanka, conquering the entire island, of which his father had occupied only the northern portion. The next year he reconquered the Pandya king to his south and the raja of Kerala to his west. In 1021 he attacked the Chalukyas of Kalyana, an ascendant dynasty based in the heart of the Deccan plateau. Having defeated that house, Rajendra returned to his capital before moving his army towards Bengal, thereby continuing his clockwise digvijaya. But the inscription discloses another rationale for the expedition to Bengal. ‘This light of the Solar race [Rajendra],’ it says,

laughing at Bhagiratha who had brought down the Ganga [to the earth from heaven] by the power of [his] austerities, wished to sanctify his own country with the waters of the Ganga [i.e. the Ganges] carried thither through the strength of [his] arm. Accordingly [he] ordered the commander of the army who had powerful battalions [under his control], who was the resort of heroism [and] the foremost of diplomats – to subdue the enemy kings occupying [the country on] the banks of that [river].¹¹

This passage refers to a Hindu myth, visually narrated in a stunning seventh-century seaside bas-relief at Mahabalipuram (or Mamallapuram, near modern Chennai), according to which the ascetic Bhagiratha, by performing rigorous austerities, induced the great god Śiva to water the parched earth by bringing the Ganges down from heaven. The parallel between Bhagiratha and Rajendra is clear: if an ascetic had mythically brought the river down from heaven to earth, King Rajendra would ritually bring it down from north India to Tanjavur.

Rajendra attached great importance to his raid on Bengal and the pots of Ganges water that he brought south to his capital. Not only did he assume the title Gangaikonda-Chola (‘the Chola who took the Ganges River’), but he built a new capital in the Kaveri delta named Gangaikonda Cholapuram, or ‘the city of the Chola who took the Ganges’. This he embellished with a colossal temple to Śiva whose central, nine-storey shrine soars to a height of fifty-six metres. Inside, he had a well dug for the sacred